← 返回列表

Reproduced Vulnerabilities

单站 cve.imfht.com 发布 2026-11-18 入库 2026-11-18T09:02:38

Every card below is a CVE our Claude Code agent reproduced end-to-end: it reads the PoC, rebuilds the real vulnerable system in an isolated Docker sandbox, launches a real exploit, and records the whole run with asciinema. A "VULNERABLE:" line is hard proof the exploit fired.

Full sandbox recordings + exploit POC are a Pro+ exclusive.
Upgrade to Pro+ — limited ¥499/mo

File Thingie 2.5.7 Authenticated Arbitrary File Upload Remote Code Execution

Success marker:
VULNERABLE: zip-upload/unzip bypass -> PHP webshell RCE confirmed, exfiltrated PROOF_d735075b52968c27 from /flag.txt, uid=33(www-data) gid=33(www-data) groups=33(www-data)

Tenda AC20 httpd setPptpUserList formSetPPTPUserList stack-based overflow

pdfminer.six vulnerable to Arbitrary Code Execution via Crafted PDF Input

Success marker:
VULNERABLE: pdfminer.six executed attacker-controlled pickle via PDF-specified CMap path traversal (uid=0(root) gid=0(root) groups=0(root)); exfiltrated PROOF_8b92c89481efabf3

LangChain serialization injection vulnerability enables secret extraction

MySQLSelectTool Read-Only Bypass via SELECT INTO OUTFILE Allows Arbitrary File Write

Success marker:
VULNERABLE: SELECT INTO OUTFILE bypass wrote an arbitrary file on the MySQL server; retrieved file content: PROOF_03f0e21ac530efe6

Roo Code is Vulnerable to Potential Remote Code Execution via zsh Command Validation Bug

Suricata is vulnerable to a stack overflow on larger compressed data

Success marker:
VULNERABLE: suricata crashed with SIGSEGV (exit code 139) while stack-allocating a ~16MB VLA for SWF/LZMA compressed data in FileSwfDecompression (CVE-2025-64332)

glob CLI: Command injection via -c/--cmd executes matches with shell:true

Success marker:
VULNERABLE: glob CLI -c command injection executed sh payload from malicious filename poc$(cat flag>stolen)END.txt; exfiltrated secret=PROOF_41b80aa21f5dcc33

code-projects Simple Food Ordering System editproduct.php unrestricted upload

Success marker:
VULNERABLE: unrestricted upload -> PHP webshell RCE as 'SHELL_OK:uid=33(www-data) gid=33(www-data) groups=33(www-data)'; /flag.txt read via shell: PROOF_d58e4cd918af61fc

campcodes Online Student Enrollment System register.php unrestricted upload

PhotoShow 3.0 Remote Code Execution via Exiftran Path Injection

Use after free in fluidsynth

flatted: Prototype Pollution via parse()

Success marker:
VULNERABLE: flatted.parse() leaked a live Array.prototype reference and a single write polluted every array - fresh [].FLAG == PROOF_e3512103754b6995

NanoCMS 0.4 - Remote Code Execution (RCE) (Authenticated)

arduino-TuyaOpen DnsServer Heap-Based Buffer Overflow Remote Code Execution

Intake has a Command Injection via shell() Expansion in Parameter Defaults

Success marker:
VULNERABLE: shell() parameter default executed as uid=0(root) inside intake 2.0.8; flag exfiltrated via injected command: PROOF_326394275fe4632f

Belkin F9K1122 Parameter formCrossBandSwitch stack-based overflow

Belkin F9K1122 Setting formSetSystemSettings stack-based overflow

Immutable.js: Improperly Controlled Modification of Object Prototype Attributes ('Prototype Pollution') in immutable

Pillow has an out-of-bounds write when loading PSD images

Success marker:
VULNERABLE: out-of-bounds write confirmed — crafted PSD psd-oob-write.psd caused SIGSEGV (exit code 139) in Pillow 12.1.0 raw decoder via unvalidated negative tile extents

MiniFtp parseconf_load_setting Buffer Overflow via Configuration

Handlebars.js has JavaScript Injection via AST Type Confusion when passing an object as dynamic partial

lodash vulnerable to Code Injection via `_.template` imports key names

Glances Vulnerable to Command Injection via Dynamic Configuration Values

parseusbs \< 1.9 Command Injection via Volume Path Argument

Success marker:
VULNERABLE: OS command injection in parseUSBs 1.8 -v flag; injected command executed as uid=0(root) and exfiltrated /flag.txt -> PROOF_cee9a8151d42428a

dynaconf Affected by Remote Code Execution (RCE) via Insecure Template Evaluation in @jinja Resolver

Success marker:
VULNERABLE: dynaconf 3.2.11 @jinja SSTI RCE — popen ran inside the target, exfiltrated PROOF_a6d4b95590d9b61e and command output uid=0(root)

xmldom: XML injection via unsafe CDATA serialization allows attacker-controlled markup insertion

Success marker:
VULNERABLE: CDATA terminator injection confirmed — serialized output contained attacker markup that re-parsed into a live <injected> element carrying PROOF_e0800ddd49c4d2a9

jackying H-ui.admin preview.php unrestricted upload

Success marker:
VULNERABLE: unauth RCE via preview.php unrestricted upload — webshell executed as uid=33(www-data), exfiltrated flag: PROOF_277d23f2a06160f5

PromtEngineer localGPT server.py do_POST unrestricted upload

pyOpenSSL DTLS cookie callback buffer overflow

BrowserStack Runner 0.9.5 Unauthenticated RCE via /_log HTTP Handler

Hermes WebUI \< 0.51.311 RCE via Git Configuration Injection

Success marker:
VULNERABLE: core.fsmonitor RCE executed as [uid=1(daemon) gid=1(daemon) groups=1(daemon)]; exfiltrated /flag.txt = PROOF_3444424284e95868

xmldom: XML node injection through unvalidated comment serialization

Success marker:
VULNERABLE: comment-breaking sequence --> injected <steal>PROOF_cc4b2731acdab1a2</steal> into XMLSerializer output

xmldom: XML injection through unvalidated DocumentType serialization

JunoClaw: plugin-shell shell-metacharacter injection via shell wrapper

Velocity.js: Prototype Pollution in #set path assignment

RVF: Prototype pollution in @rvf/set-get reachable via @rvf/core preprocessFormData (HTTP form data)

Prototype Pollution in parse-nested-form-data via `__proto__` in FormData field names

deepobj: Improperly Controlled Modification of Object Prototype Attributes ('Prototype Pollution')

Prototype pollution in form-data-objectizer via bracket-notation form keys

PhpWeasyPrint: shell command injection via configurable WeasyPrint binary path due to inverted is_executable() guard (mirror of KnpLabs/snappy GHSA-vpr4-p6fq-85jc)

jsondiffpatch 安全漏洞

Success marker:
VULNERABLE: jsondiffpatch.patch() polluted Object.prototype — a fresh {} now inherits pwned=PROOF_e5638d2a73be1724

picklescan - Arbitrary Code Execution via Undetected idlelib.pyshell.ModifiedInterpreter.runcommand

React Router's vendored turbo-stream v2 allows arbitrary constructor invocation via TYPE_ERROR deserialization leading to Unauth RCE

GitPython: Unsafe option check validates multi_options before shlex.split transforms it

AI Tensor Engine for ROCm (AITER) 0.1.14 Unauthenticated RCE via MessageQueue.recv() Pickle Deserialization

PhpWeasyPrint vulnerable to PHAR deserialization via output filename (CVE-2023-28115 case-insensitive bypass)

piscina: Prototype Pollution Gadget → RCE via inherited options.filename

shell-quote `quote()` does not validate object-token shapes, allowing command injection via line terminators in `.op`

Success marker:
VULNERABLE: unescaped LF in .op left a raw newline in quote() output; /bin/sh executed a second attacker command (cat /flag.txt) that returned PROOF_9b9d3050aad4c25b

GitPython: Newline injection in config_writer().set_value() enables RCE via core.hooksPath

pam_usb: PINENTRY_FALLBACK_APP environment variable allows arbitrary command execution

OS Command Injection in Microsoft UFO Shell Action Replay via Stored Session JSON

protobufjs: Code injection through bytes field defaults in generated toObject code

Success marker:
VULNERABLE: exfiltrated /flag.txt = PROOF_212762902e33ede6, injected code ran as uid=0(root)

Banks: Critical Remote Code Execution (RCE) via Jinja2 SSTI

Success marker:
VULNERABLE: banks SSTI -> RCE confirmed: uid=0(root), exfiltrated flag: PROOF_9644a9af85a9fa39

Integer Overflow in rxi/microtar mtar_next() Causes Infinite Loop DoS

zhayujie chatgpt-on-wechat Bash Tool bash.py _get_safety_warning os command injection

Success marker:
VULNERABLE: os command injection via chatgpt-on-wechat Bash tool guard bypass - shell ran as uid=0(root), exfiltrated protected credential: PROOF_33dafa5767c82481

LangSmith Client SDK: Public prompt pull deserializes untrusted manifests without trust boundary warning

Incomplete mitigation of CVE-2026-4519, %action expansion for command injection to webbrowser.open()

Picklescan - Arbitrary Code Execution via numpy.f2py.crackfortran._eval_length Gadget

picklescan - Remote Code Execution via Undetected profile.Profile.runctx

picklescan - Arbitrary Code Execution via Undetected ensurepip._run_pip Function

Success marker:
VULNERABLE: ensurepip._run_pip bypassed picklescan 0.0.26; arbitrary code executed as 'uid=0(root) gid=0(root) groups=0(root)'; exfiltrated secret=PROOF_4c8cb37d6823c6a8

picklescan - Arbitrary Code Execution via Undetected trace.Trace.run in Pickle Files

picklescan - Remote Code Execution via idlelib.debugobj.ObjectTreeItem.SetText

Success marker:
VULNERABLE: picklescan 0.0.28 passed the malicious pickle and pickle.load() executed embedded code as uid=0(root), exfiltrating flag=PROOF_71a5c75515fd2eae

picklescan - Remote Code Execution via idlelib.autocomplete.AutoComplete.get_entity

picklescan - Arbitrary Code Execution via Undetected cProfile.run in Pickle Deserialization

Success marker:
VULNERABLE: picklescan 0.0.29 scan passed (exit 0, no alarm) while pickle.load executed cProfile.run('cat /flag.txt') retrieving secret: PROOF_ca90cfa80c847167

picklescan - Arbitrary Code Execution via numpy.f2py.crackfortran.myeval Detection Bypass

picklescan - Remote Code Execution via code.InteractiveInterpreter Detection Bypass

picklescan - Arbitrary Code Execution via Undetected profile.Profile.run

picklescan - Arbitrary Code Execution via Undetected idlelib.autocomplete.AutoComplete.fetch_completions

picklescan - Remote Code Execution via Undetected cProfile.runctx in Pickle Files

Success marker:
VULNERABLE: picklescan 0.0.29 reported 0 dangerous globals (bypass confirmed) and pickle.load() executed cProfile.runctx payload -> exfiltrated flag: PROOF_a462898f3cd65e49

MessagePack: Out-of-bounds read/crash on Unpacker reuse after caught error

Success marker:
VULNERABLE: SEGV confirmed — msgpack 1.2.0 Unpacker crashed with SIGSEGV (exit code 139) on reuse after caught error (OOB read)

Laravel-Mediable \< 7.0.0 File Upload RCE via Extension Bypass

Success marker:
VULNERABLE: stored shell.php.jpg executed as PHP code (double-extension bypass) — ����JFIF��uid=33(www-data) gid=33(www-data) groups=33(www-data) PROOF_1ccffff04d2994e8

Authenticated terminal command whitelist bypass in Pheditor

`datamodel-code-generator` vulnerable to code injection in via attacker-controlled `default_factory` schema field

CedarJava has policy injection, type confusion, and incorrect equality comparison vulnerabilities

Success marker:
VULNERABLE: EntityIdentifier.equals() inverted at runtime (null-equals:true, self-equals:false); null-token auth bypass exfiltrated secret PROOF_55965fa58f9d0963

CedarJava has a policy injection vulnerability

PraisonAI before 4.6.78 Allowlist Bypass via find -exec

9Router: Authenticated RCE via Unvalidated MCP Plugin Arguments

GitPython before 3.1.51 Command Injection via option prefix abbreviation

Success marker:
VULNERABLE: git executed attacker command via abbreviated --upload-p option (blocklist exact-match bypass); exfiltrated /flag.txt content: PROOF_69ef3d1150e167ea

Incomplete fix for CVE-2026-25754 in @adonisjs/bodyparser

PraisonAI before 1.6.78 Remote Code Execution via Plugin Auto-Discovery

picklescan - Undetected Remote Code Execution via idlelib.run.Executive.runcode

Success marker:
VULNERABLE: picklescan<0.0.30 passed the malicious idlelib.run.Executive.runcode pickle as clean; pickle.load() executed it -> RCE as uid=0 | flag=PROOF_c4723c4b9ebb3419

picklescan - Arbitrary Code Execution via lib2to3.pgen2.pgen.ParserGenerator.make_label Detection Bypass

picklescan - Unsafe Deserialization via lib2to3.pgen2.grammar.Grammar.loads

picklescan - Remote Code Execution via Undetected idlelib.calltip.get_entity

picklescan - Arbitrary Code Execution via Unsafe Deserialization in numpy.f2py.crackfortran

picklescan - Remote Code Execution via _operator.attrgetter Detection Bypass

picklescan - Undetected Remote Code Execution via _operator.methodcaller

Savon::Model evaluates WSDL operation names as Ruby source

NVIDIA Megatron Bridge 代码注入漏洞

`datamodel-code-generator` vulnerable to code execution on import via unescaped `validators` entries in --extra-template-data

Success marker:
VULNERABLE: injected validator executed on model import — /flag.txt read via code execution, token: PROOF_a2e5f896bbb2ed7a

Wazuh: Heap buffer overflow in wazuh-analysisd via rootcheck event parsing

Success marker:
VULNERABLE: wazuh-analysisd was running before the event and is GONE after it - heap overflow in W_JSON_ParseRootcheck killed the SIEM analysis engine

datamodel-code-generator vulnerable to code injection via `x-python-import` / `customTypePath` in generated import statements

Success marker:
VULNERABLE: CVE-2026-55415 code injection executed on model import — flag exfiltrated via generated code: PWNED:PROOF_814b7610a98b8ca2

Kiota: Code Generation Literal Injection in the Python Generator

Success marker:
VULNERABLE: attacker-controlled x-ms-enum description executed as module-scope Python on import — proof token from generated code: PROOF_edade13faf67dd41

Banks: Unsafe importlib.import_module of attacker-controlled Tool.import_path in CompletionExtension allows RCE

Success marker:
VULNERABLE: banks 2.4.2 executed attacker-controlled import_path=subprocess.getoutput in-process — uid=0(root), /flag.txt exfiltrated through the RCE: PROOF_b56db402aabcbd16

@phun-ky/defaults-deep Has a Prototype Pollution issue via Unsafe Recursive Property Merging

ArcadeDB before 26.7.2 Remote Code Execution via Trigger Scripts

DragonflyDB `CMS.INITBYDIM` integer overflow leads to a remote, attacker-controlled heap out-of-bounds write

backmeup (npm) - OS Command Injection via Backup Option Values

Success marker:
VULNERABLE: OS command injection confirmed — injected shell command ran (uid=0(root) gid=0(root) groups=0(root)) and exfiltrated PROOF_50c77c9f9b80f15a from /flag.txt

OpenChoreo: Authenticated OS command injection via OpenChoreo Workflow Plane templates enables code execution in privileged pods

rsync \< 3.5.0 Command Injection via Multiple Code Paths

compliance-trestle Vulnerable to Remote Code Execution via Recursive Server-Side Template Injection (SSTI)

Success marker:
VULNERABLE: RCE via recursive SSTI in trestle author jinja - payload executed as uid=0(root), exfiltrated /flag.txt = PROOF_86516fb7ee369a23

Oh My Posh: Arbitrary command execution via template injection in the path segment

Trippo ResponsiveFilemanager dialog.php unrestricted upload

Glances: `--disable-config-exec` does not cover on-alert action commands (incomplete fix of CVE-2026-53925)

mcp-shell — Security Disabled by Default in Bare-Binary Deploy Path + Shell Interpreter in Secure-Mode Allowlist

bestzip before 2.2.6 and 3.0.x before 3.0.2 Argument Injection via Missing Option Delimiter

Success marker:
VULNERABLE: bestzip 2.2.5 argument injection - zip ran the attacker "-TT" command via shell and exfiltrated /flag.txt: PROOF_b08c6b2592ad34d6

GitPython before 3.1.58 Arbitrary File Overwrite via read-tree

Argo-cd: argo cd unauthenticated remote code execution in repo-server via generatemanifest grpc endpoint

Success marker:
VULNERABLE: unauthenticated repo-server GPRC GenerateManifest RCE + manifest cache injection confirmed; attacker-controlled manifest content retrieved: PROOF_e62c8f3312788169

Penpot: MCP REPL server binds to 0.0.0.0 with unauthenticated /execute endpoint — RCE

Success marker:
VULNERABLE: unauthenticated /execute on 0.0.0.0:4403 executed arbitrary JS as uid=0 flag=PROOF_0730922deaa80032 and exfiltrated proof token from /flag.txt

Pi-hole: Local privilege escalation from `pihole` user to root via `/etc/pihole/logrotate`

genieacs-mcp: DNS rebinding reaches local GenieACS MCP Streamable HTTP transport

Kiota: Code Generation Literal Injection in the PHP Generator

browse-mcp: Arbitrary file write via unconfined download and state paths

Success marker:
VULNERABLE: arbitrary file write outside download root — token PROOF_6146d922968de4be recovered from /tmp/steal.txt (browser_download save_dir=/tmp honored an absolute path)

PraisonAI before 1.6.78 DNS Rebinding SSRF via web_crawl

Success marker:
VULNERABLE: web_crawl SSRF bypassed hostname guard and exfiltrated internal loopback response body with proof token PROOF_02ceec9ed189ee52

Dash: Users can write to config despire permissions (OIDC tested)

Success marker:
VULNERABLE: unauthenticated user overwrote /app/user-data/conf.yml despite OIDC auth + admin permissions being configured (title is now EXPLOITED_BY_UNAUTHENTICATED_USER)

DOMPurify XSS via `selectedcontent` re-clone

MCP Appium: Unescaped Locator Data XSS in MCP-UI Resource (createLocatorGeneratorUI)

hbs vulnerable to XSS via registerAsyncHelper output-escaping bypass

Success marker:
VULNERABLE: XSS confirmed - raw <script>alert(1)</script> injected unescaped into server-rendered HTML via hbs registerAsyncHelper (CVE-2026-16231)

NVIDIA NemoClaw 授权问题漏洞

Success marker:
VULNERABLE: SSRF confirmed — 0.0.0.0/8 endpoint passed validateEndpointUrl() v0.0.12 (missing 0.0.0.0/8 block) and internal secret exfiltrated: PROOF_ac84660f96cfc1f4

Twig: `template_from_string()` escapes a SourcePolicy-driven sandbox via synthesized template name

Directus: SSRF Protection Bypass via 0.0.0.0 in File Import

Puma PROXY Protocol v1 Parser Allows Remote Memory Exhaustion

Success marker:
VULNERABLE: unbounded PROXY v1 buffer growth confirmed (RSS grew 7488 KB after 8MB no-CRLF stream)

Soup Sieve: Memory Exhaustion via Large Comma-Separated Selector Lists in soupsieve

Success marker:
VULNERABLE: 1999999-byte selector list allocated 469 MB of heap (peak 483 MB RSS) via unbounded comma-separated list in soupsieve.compile(); proof token PROOF_f7418e3960b2f15f

Pillow: Heap out-of-bounds write `Image.paste()` / `Image.crop()` via signed coordinate overflow

Success marker:
VULNERABLE: Image.paste() OOB heap write confirmed - process crashed with exit code 139 (SIGSEGV); retrieved token PROOF_ec1b6a798dd202d2 from container

Pillow: Decompression Bomb DoS via PdfParser.PdfStream.decode()

Success marker:
VULNERABLE: MemoryError (exit code 1) from unbounded zlib.decompress in PdfParser.PdfStream.decode(); proof token PROOF_39a713bd6a6a8fe3 confirmed: PROOF_39a713bd6a6a8fe3

Pillow: Controlled heap out-of-bounds write in `ImageCmsTransform.apply()` via output mode mismatch

Success marker:
VULNERABLE: ImageCmsTransform.apply() accepted mismatched output image mode (no ValueError raised, C code executed with wrong stride) - fixed in 12.3.0 by mode check

pyasn1 BER/CER/DER decoder denial of service via unbounded long-form tag IDs

pyasn1: Quadratic complexity in OBJECT IDENTIFIER and RELATIVE-OID processing allows denial of service

Success marker:
VULNERABLE: quadratic OID decode, 40000 arcs took 2504ms (5k=37ms -> 40k=2672ms, ~40x for 8x input = quadratic), proof flag retrieved: PROOF_7ffc9670f57590ec

PraisonAI before 4.6.78 Unenforced Security Policy in Subprocess Sandbox

luci-app-banip Log Monitor IP Extraction Bypass

Success marker:
VULNERABLE: banIP blocked attacker-injected IP 10.0.66.66 (from username field) instead of real attacker 203.0.113.99; ban set = [10.0.66.66]

code-projects Online Job Portal DeleteUser.php sql injection

code-projects Online Job Portal JobSeekerInsert.php unrestricted upload

Success marker:
VULNERABLE: unrestricted upload via txtFile led to RCE - SHELL_EXEC_OK uid=www-dataPROOF_7bcd3d50576c283e (exfiltrated PROOF_7bcd3d50576c283e from /flag.txt)

PraisonAI before 4.6.78 Authentication Bypass via HTTP-stream

Success marker:
VULNERABLE: unauthenticated MCP tools/call (no Authorization/Origin header) read secret /flag.txt via praisonai.workflow.show -> exfiltrated PROOF_9a9c48382e73788b

Faktory: Insecure predictable /tmp/redis.conf enables local Redis config hijack (network exposure / root RCE primitive)

Success marker:
VULNERABLE: attacker-planted /tmp/redis.conf hijacked embedded redis (unauthenticated PONG on 0.0.0.0:9999); exfiltrated proof flag: PROOF_7ef512bac4e91072

cleverbrush framework/deep deepExtend.ts deepExtend prototype pollution

Success marker:
VULNERABLE: 33[0mVULNERABLE: prototype pollution via deepExtend — Object.prototype.pwned = PROOF_e950868b3d6b5ae3

Kiota: Generation-time SSRF + remote/local file inclusion via unrestricted $ref

Kiota: Workspace-config poisoning: out-of-repo file write + generation-time SSRF

NVIDIA NemoClaw 代码注入漏洞

Success marker:
VULNERABLE: prototype pollution in NemoClaw migration command; exfiltrated host secret via polluted prototype: PROOF_0f3e5c789a484cc9

Tornado before 6.3.3 HTTP Request Smuggling via Content-Length

Success marker:
VULNERABLE: Tornado accepted Content-Length with embedded CRLF and executed the smuggled request: token PROOF_5fd4646febd71897 returned from second (smuggled) request

Netty: DNS Codec Input Validation Bypass in Netty (Encoder + Decoder)

js-yaml quadratic-complexity denial of service via YAML11_SCHEMA !!omap parsing

Success marker:
VULNERABLE: CVE-2026-59870 RESULT:VULNERABLE 5.2.0 n=20000 elapsed=6238ms n=10000 elapsed=799ms ratio=7.81 (quadratic growth in !!omap parsing)

CodeIgniter validation of header name and value

Agentic-Flow: OS Command Injection in agentic-flow MCP server tools via unsanitized tool-parameter interpolation into execSync

@fastify/reply-from vulnerable to cross-upstream request routing via URL cache key collision

Docling Core has unsafe remote filename resolution

Success marker:
VULNERABLE: server-controlled Content-Disposition made docling-core write remote content to /flag.txt (secret exfiltrated: PROOF_f7d087fd0992b332)

Gestor de Oferta: Prototype pollution in @tmlmobilidade/utils setValueAtPath

Success marker:
VULNERABLE: prototype pollution confirmed — ({}) inherited polluted property, exfiltrated via /tmp/polluted.json: {"leaked":"PROOF_7c4a8c024da77542"}

Docling Core has insufficient validation of image reference URIs

Pimcore: Unsafe PHP Deserialization in Multiple Locations Without allowed_classes Restriction

Success marker:
VULNERABLE: unsafe unserialize() -> __destruct RCE in Pimcore affected source: RCE as uid 0 root ; proof token exfiltrated: PROOF_02e5ef68250e1b19

Gerapy Project Upload Endpoint views.py missing authentication

Success marker:
VULNERABLE: unauthenticated POST /api/project/upload (no auth token sent) wrote /flag.txt via zip-slip; leaked PROOF_b7c9e5be176561cc

oras-go: Hardlink entry with relative Linkname escapes extract dir via process CWD resolution in `oras-go` tar extraction

GitPython before 3.1.54 Remote Code Execution via kwarg value smuggling

Success marker:
VULNERABLE: smuggled --upload-pack executed during clone; exfiltrated PROOF_3b6c16cf5c799663 from /flag.txt via git

wetty vulnerable to DOM XSS via file-download filename

SVGO: removeScripts plugin leaves some executable scripts intact

Tenda CH/CP/TX3 SSH hard-coded password

electerm: Path traversal in FTP/SFTP recursive folder download via unsanitized server filename

GitPython before 3.1.57 Arbitrary File Overwrite and Read

Success marker:
VULNERABLE: arbitrary file read via TagReference.create(-F) returned in-band secret: PROOF_57b6936950faf615

rsync \< 3.5.0 Symlink Following Arbitrary File Overwrite

Flyto2 Core \< 2.28.0 SSRF Guard Bypass via is_private_ip()

Hub: Unauthenticated WebSocket RPC Waiter Resource Exhaustion

Success marker:
VULNERABLE: resource exhaustion — RSS grew 6868 kB (57448 -> 64316 kB) from leaked RPC waitForReply timers after 250 silent connections; container token PROOF_1b4d4eb452818b8b

js-yaml: Exponential parsing time in the flow collections leads to denial of service

FreePBX: Authenticated TTS AGI Command Injection Through TTS Name

Success marker:
VULNERABLE: TTS AGI command injection - uid=0(root), injected \$( ) executed, flag PROOF_70bafb44b0e1068f present in container

compliance-trestle Remote Fetching Mechanism has an Arbitrary File Write via Cache Path Traversal

Success marker:
VULNERABLE: arbitrary file write confirmed - PROOF_c5dd6612da330b8b (content retrieved: PROOF_c5dd6612da330b8b)

WsgiDAV encoded dot segments can escape filesystem share roots

Success marker:
VULNERABLE: path traversal via encoded dot segment - served /var/wsgidav-root/share-evil/flag.txt outside share root (HTTP/1.1 200 OK); content: PROOF_a25ba7daaac9191d

RabbitMQ Java client ValueReader: Unbounded recursive table/array nesting causes StackOverflowError DoS

Success marker:
VULNERABLE: StackOverflowError from unbounded recursion in com.rabbitmq.client.impl.ValueReader.readTable; proof token PROOF_395418ef20200bf1 exfiltrated

goshs: WebDAV listener ignores --read-only, --upload-only, and --no-delete mode flags

Success marker:
VULNERABLE: WebDAV PUT succeeded despite --read-only; file on disk after PUT: "PWNED" (webdav port ignores mode flags, pre-2.1.0)

Actor MCP path authority injection leaks Apify token

Success marker:
VULNERABLE: victim Apify token Bearer apify-token-SECRET-VICTIM-TOKEN-4c4a4030e1f79e1e exfiltrated to attacker MCP host (flag: PROOF_4c4a4030e1f79e1e)

Dokploy: Invitation Role Escalation Allows Organization Takeover

sh _uid does not drop supplementary groups (incomplete privilege drop)

Success marker:
VULNERABLE: sh 2.2.3 _uid=64001 child RETAINED root supplementary group (groups=64001) and read /flag.txt -> PROOF_ded98122a3224b6d

BetterDesk has a replay behavior vulnerability when devices are deleted

Mailu missing authentication on PATCH /api/v1/token/\, which allows unauthenticated removal of IP restrictions

SourceCodester Class and Exam Timetabling System edit_subject1.php sql injection

Success marker:
VULNERABLE: SQLi via edit_subject1.php?id= — exfiltrated DB secret through UNION: PROOF_19a14d1b7a019acc

PHPGurukul Complaint Management System check_availability.php sql injection

Gotenberg: Path traversal in zip entry name via Windows-style separators in upload filename

Arc: Unauthenticated access to Go debug pprof endpoints leaks runtime state and enables CPU-burn DoS

4gaBoards: Pre-Account Takeover via SSO Email Linkage

GitPython before 3.1.58 Command Execution via split_single_char_options

Success marker:
VULNERABLE: RCE via check_unsafe_options bypass, exfiltrated flag: PROOF_4b7d031af1c3bff7

GitPython before 3.1.58 Config Injection via option-name

PraisonAI serve agents --api-key is ignored, allowing unauthenticated remote agent execution

Success marker:
VULNERABLE: unauthenticated POST to /agents/proof_agent returned 200 with agent-executed proof token PROOF_ec71df256b496f69 despite --api-key being set — api key is ignored

stigmem before 0.9.0a12 Cross-Tenant BOLA via quarantine

Success marker:
VULNERABLE: cross-tenant BOLA — tenant-A key listed tenant-B quarantined fact containing proof token PROOF_ed5a7b3721de5dbf

Grav: .htaccess file extension rules bypass via case variation on case-insensitive filesystems

GitPython before 3.1.58 Path Traversal via .gitmodules Submodule Name

Missing JSON Web Token signature validation in Otalio Ship Property Management System

Success marker:
VULNERABLE: JWT with INVALID signature accepted as admin; exfiltrated secret PROOF_56975c534318ae9e from /api/admin/ship-records — privilege escalation confirmed

BabelDOC: Arbitrary Code Execution via CMap Pickle Deserialization in babeldoc/pdfminer/cmapdb.py

Success marker:
VULNERABLE: arbitrary code execution via CMap pickle deserialization — payload ran as the BabelDOC process, exfiltrated proof token: PROOF_0826adf5d4c4997d

4gaBoards: SSO Pre-Account Takeover / Hijacking via Mass Assignment

Gotenberg: SSRF deny-list bypass in IsPublicIP via IPv6 6to4 / NAT64 / site-local prefixes

GitPython before 3.1.58 Remote Code Execution via Repo.init

Success marker:
VULNERABLE: RCE via Repo.init(template=...) - malicious post-commit hook executed as uid=0(root) gid=0(root) groups=0(root), exfiltrated token from /flag.txt

libcrux before 0.0.6 Cryptographic Implementation Bug Fixes

Success marker:
VULNERABLE: libcrux-ecdh 0.0.5 X25519 secret validation bug confirmed - unclamped and wrong-length secrets accepted, PROOF_f4055944652a194e

stigmem Postgres SQL Injection via Schema Identifier

Success marker:
VULNERABLE: SQL injection via unquoted Postgres schema identifier — injected DDL executed in node database, exfiltrated proof token PROOF_bb47224dc3e89295 from sqli_proof

SourceCodester Simple Online Food Ordering System ajax.php save_menu sql injection

stigmem Plugin Signature Enforcement Bypass via Configuration

Success marker:
VULNERABLE: unsigned plugin code executed via single-flag signature bypass (STIGMEM_PLUGIN_SIGNING_REQUIRED=false, no ack); exfiltrated per-run token: PROOF_7f0bd8f025748e02

SourceCodester Simple Online Food Ordering System view_prod.php sql injection

Success marker:
VULNERABLE: SQL injection in /fos/view_prod.php?id= — exfiltrated secret token PROOF_4ce5e0ad169d1358 from DB via UNION

itsourcecode Payroll System admin_class.php save_settings unrestricted upload

Success marker:
VULNERABLE: RCE confirmed via unrestricted PHP upload — exfiltrated flag: PROOF_d8e4da7356aed211

stigmem before 0.9.0a12 Cross-Tenant BOLA via Tombstones

Arc has an authenticated arbitrary local-file read via DuckDB I/O functions that bypasses RBAC table-level checks

Success marker:
VULNERABLE: read_text(/app/flag.txt) exfiltrated PROOF_2a4fc64b9f209594 from the container filesystem via /api/v1/query

Omnigent Guardrail policy bypass: shell-command parser fails open in policies/builtins/_shell.py

SSHFS: Improper Neutralization of Argument Delimiters in a Command ('Argument Injection')

Success marker:
VULNERABLE: PROOF_ed9649253ebd090f uid=0(root) gid=0(root) groups=0(root)

原文链接:https://cve.imfht.com/reproduced